Summary
At the end of last month, the Government published the “Data Centre Covenant”1 setting out “new tests all data centres are encouraged to meet, and which strategic data centres must meet to qualify for additional action.”
It was released in parallel to the Government’s response to its consultation on “Accelerating electricity for network connections for strategic demand” (the Connections Consultation Response)2 and the release of the Initial Proposal for a National Policy Statement (NPS) for Data Centres (the draft NPS)3, the consultation for which closes on 8 January 2027.
Together, they reinforce the Government’s position on the need for data centres, update on how to mitigate the consequent impacts, and secure more local/regional benefits.
As ever, achieving balance is not straight-forward and there is always a risk that imposing measures/restraints on development will end up deterring investment and compromising the ability to achieve the highlighted macro national interest/security ambitions arising from such development. This balancing exercise will no doubt be seen as even more politically important now following the mid-term results in the US and some of the political/constituent reaction to AI/data centres.4
In this context, we explore below what the Data Centre Covenant says, some of the related opportunities and challenges for developers in complying with its tests, and how it is proposed to be translated into planning policy under the draft NPS. Whilst the NPS (once in force) will only have direct effect in relation to the data centre applications directed into the NSIP regime, it is still capable of forming a material consideration in decision making on applications that fall under the Town and Country Planning Act 1990.
The draft NPS proposes a “presumption of consent”. There are certain conditions and exceptions to this principle, but the clear benefit of this presumption and its framing in policy is that even where residual impacts from the scheme are identified – “in all but exceptional circumstances, it is unlikely that consent will be refused on the basis of these residual impacts.”
This represents powerful policy support for data centre DCO applications. In circumstances where data centres do become more contentious at a local political level, making planning less certain (both in terms of the consent, and the timescale needed to obtain it), this comparatively definitive policy position under the draft NPS and the generally reliable timescales under the DCO regime may prove attractive to developers considering their consenting route.
What does the Data Centre Covenant say?
The Covenant sets out five specific tests that it states all data centres are “encouraged” to meet, and which strategic data centres must meet to “qualify for additional government action”, specifically commitments to:
- Lower electricity bills. Data centres are to “pay for their fair share of energy – covering any energy costs they create – so households, communities and other businesses do not pick up the bill”. It is not expressly referred to in this section of the Covenant, but it is assumed that this (at least in part) anticipates the adoption of the Data Centre Commitment fee, which has been subject to recent NESO consultation.5 If it’s not solely linked to the proposed Commitment fee, then it begs the question as to how they’re otherwise anticipating this “test” to be satisfied.
- Invest in improving local communities. This is familiar policy rhetoric from this Government (and indeed its recent predecessors), with a repeat of the direction that “communities that host nationally important infrastructure should share directly in the rewards”. To do so, data centre developers are to agree a "Local Benefit Plan" that “commits substantial investment, [is] developed with the local community and linked to the development’s size.” This is separate from/additional to the need for any obligations under a planning agreement to mitigate the impact of the scheme, or community infrastructure levy (CIL) requirements. The level of such investment is not specified, but the recently published Community Benefits Factsheet alongside the Energy Independence Bill sets an expectation of £5,000 per MW per year for eligible onshore wind projects, and £700 per MW per year for eligible ground mount solar projects, which may be instructive (and indeed highlights that some care will be needed in the framing of the requirement to avoid any double-payment where the data centre is co-located with renewables generation that has a similar community benefit expectation).6 Hopefully there will be further consultation on what is envisaged by way of the Local Benefit Plan, before any such direction is imposed.
- Adhere to the highest environmental standards. Developments are to “protect nature, minimise water use and publish annual energy and water consumption reports”. As part of this, the Covenant continues by saying “they should also capture the heat they produce for nearby homes, schools, hospitals or businesses wherever practical, rather than letting it go to waste.” We discuss the translation of this “test” into the draft NPS further below, and some of the practical challenges that may result if care is not taken on the policy expectation here.
- Create jobs, apprenticeships and drive reindustrialisation. Developments should “Create local jobs and invest in paid apprenticeships and training places, supporting young people into the AI, digital and infrastructure jobs of the future. They should also prioritise UK suppliers and technologies and make a meaningful contribution to the development and deployment of British AI hardware.” These are clearly key pillars of the Government’s support for data centres, and are complementary to the final security/sovereignty ‘test’ discussed below.
- Strengthen Britain’s security. Whilst pitched as a "test", in many ways this is the fundamental “need” for such development in the Government’s mind, noting such developments are necessary to “provide secure, resilient access to compute for critical services and national priorities”. The introductory paragraphs to the Covenant also elaborate on the Government’s nationalist/sovereignty focus here, noting that “to cement our position in the new world of AI and protect ourselves from the risks, we must not allow our use of AI data centres to be reliant almost entirely on other countries. As AI and data become integral to the NHS, defence and other critical services, the UK cannot rely entirely on infrastructure located overseas. Domestic data-centre capacity will give us greater resilience, and more options to protect our national interests if global compute supply becomes constrained.”
The Covenant notes that these tests will be kept under review and updated as necessary “to ensure that commitments remain proportionate and deliver clear value to communities and the wider UK”.
What opportunities does the Data Centre Covenant bring?
In return for meeting these tests, the Government states it will establish a new Ministerial Control Centre for considering supportive action for strategic data centres.
The Covenant notes that exactly what constitutes a “strategic” data centre will be made clear in further details to be published in due course, together with the “additional criteria such projects will be expected to meet”. But, for those projects that do meet such a definition, the key advantage will come in potential access to a grid connection. The Covenant notes that for such schemes, the government will: (i) reserve future capacity and reallocate capacity released when it leaves the connections queue; (ii) work with NESO to identify additional capacity; and (iii) accelerate connections before 2030 where possible.
Each of these commitments is discussed in more detail in the Connections Consultation Response, which was published on the same day as the Covenant (a cynical view would be that the Covenant was prepared to provide political balance to some of the pro-data centre development positions otherwise put forward in the Connections Consultation Response …).
The Connections Consultation Response confirms the Government’s intention to use powers under the Planning and Infrastructure Act 2025 to implement prioritisation measures for identified strategic demand projects (including data centres) through the (and ahead of the next) NESO connections gated process:
“These prioritisation measures will enable network capacity to be allocated to strategic demand projects, when it is released by projects exiting the queue. They will also allow the government to reserve future capacity for individual strategic projects that are not yet ready to join the queue, and reserve future capacity on a non-project specific basis. In addition, the measures will be used to give new, ‘ready’ strategic demand projects higher priority in queue formation as they apply in future.”
It notes the intention for the Connections Accelerator Service (CAS) to be the mechanism through which strategic demand projects are selected to potentially benefit from these prioritisation measures, and which will depend upon: (i) sponsorship from either a government department, a devolved government or a Mayoral Strategic Authority; and (ii) determination against a Strategic Framework Document, the details of which are still to be published, but which will provide details of the scope, criteria and process for selecting which projects can benefit from these new prioritisation measures through CAS.
The Connections Consultation Response notes the high level of agreement amongst respondents to the consultation on the proposed prioritisation measures, and so the Government’s confirmation of its intention to pursue these measures is not surprising. Although not an express subject of that consultation, the Response also noted the broad support for the introduction of a “Data Centre Commitment fee” “to encourage less viable projects in the existing queue to self-terminate”. The Response notes the introduction of this fee is still subject to further discussion/deliberation with NESO, but notes it is expected to be applied to projects above 40MW connecting to or having an impact on the Transmission system and to be in place before NESO reopens the process for transmission connection applications.
What are the challenges with the Data Centre Covenant?
The Government clearly hopes the aggregate effect of these measures will help unlock development and delivery of those data centre projects it has found to be strategically important.
Whilst, in isolation, these measures represent a coherent approach to resolving a/the major obstacle to data centre development (access to Grid), the Connections Consultation Response has a second key output which proposes conditionality to such development – the alignment of data centre connections after 2030 to regional infrastructure targets.
This measure anticipates “regional targets will place limits on the volumes that can connect in each region, based on assessment of grid capacity and anticipated future data centre demand, as well as wider factors including water and land availability and the needs of other types of demand.” This “spatial/zoning” approach to development and connections is consistent with the rationale informing the Government’s related Strategic Spatial Energy Plan and Centralised Strategic Energy Plan (amongst others).
This particular strand of the Consultation was more contentious, with only half of the respondents supporting the proposition to varying degrees. Further, whilst only 13% disagreed with the proposal, the summary of such disagreement provided in the Response highlights a commercial/delivery tension that I fear may be ultimately more impactful than the Government has presented in the balancing exercise:
“13% (18) of respondents disagreed with introducing strategic alignment for data centre connections. Many of these respondents perceived quotas or regional capacity pots as inflexible central planning that would add bureaucracy and constrain viable schemes. They noted that data centre siting depends on wider commercial and technical factors, so regional targets based solely on network considerations could misdirect projects to less viable locations. Some demand project developers warned that prescriptive UK rules could increase uncertainty and divert globally mobile investment to Europe or North America.”
This comment neatly encapsulates the challenge with a “top-down” approach to development planning and the approach will need to be cognisant of such risk to avoid deterring investment. The other obvious challenge for delivery that comes with seeking to focus development towards identified regional capacity is that there is an inevitable short/medium-term delay whilst the wider plan is prepared and its outputs confirmed. I also fear that the pace at which this infrastructure is required and the global competition for the necessary capital deployment isn’t naturally sympathetic to this level of careful, iterative, planning (and that’s assuming the output of the exercise once complete is fit for purpose on delivery, which is by no means guaranteed given the nature and complexity of the task).
What is the application of this to planning?
The Government also published the draft NPS on the same day as the Covenant and Connections Consultation Response.
The draft NPS proposes the policy that will inform the determination of data centre developments “directed into” the NSIP/DCO regime by the Secretary of State pursuant to requests made by developers under section 35 of the Planning Act 2008. It sets an indicative 100MW minimum threshold for data centres capable of being directed in, and notes this is intended to ensure that this route is reserved for only “the most nationally significant developments”.
Whilst we discussed above that what constitutes a “strategically important” data centre for the purposes of the Covenant and grid connection prioritisation measures remains to be confirmed, it is reasonable to assume that this 100MW threshold may be instructive as to the Government’s current thinking. It would seem incoherent if a scheme was directed into the NSIP regime as being “nationally significant”, but subsequently considered by the Government to not be “strategically important”.
The draft NPS repeats the Critical National Infrastructure designation of data centres7 and reinforces the need for such infrastructure (principally driven by the national security/sovereignty benefits highlighted in the fifth limb to the Covenant). In response, the policy statement proposes a “presumption of consent” for relevant applications which it states to mean (at paragraph 82) that provided the application accords with the requirements of the policy, consent should be granted. There are certain conditions and exceptions to this principle, but the clear benefit of this presumption and its framing in policy is that even where residual impacts are identified – “in all but exceptional circumstances, it is unlikely that consent will be refused on the basis of these residual impacts.”
This represents powerful policy support for data centre DCO applications (akin to the Critical National Priority designation for energy projects under the Overarching National Policy Statement for Energy, which has been very helpful for DCO schemes benefiting from that policy to date). In circumstances where data centres do become more contentious at a local political level, making planning less certain (both in terms of the consent, and the timescale needed to obtain it), this comparatively definitive policy under the draft NPS and the generally reliable timescales under the DCO regime may prove attractive to developers considering their consenting route. This is particularly so now that the legal duty to carry out pre-application consultation has been removed from the DCO regime by the Planning and Infrastructure Act 2025, creating greater parity with the Town and Country Planning Act regime.
Whilst the NPS (once in force) will only have direct effect in relation to the data centre applications directed into the NSIP regime, it is still capable of forming a material consideration in decision making on applications that fall under the Town and Country Planning Act 1990 (see paragraph 11 of the draft NPS) and so data centre developers, particularly of a scale exceeding 100MW, would be wise to consider their scheme against the policies and assessment criteria in the draft NPS and particularly alignment with the Data Centre Covenant tests, regardless of whether or not they are pursuing a DCO direction and subsequent application.
Referring back to the Covenant and consistent with the commonality of their release, the foreword to the draft NPS notes that “This NPS establishes robust assessment criteria, aligned with the tests set out in the government's Data Centre Covenant…”.
Specifically, in relation to “lower electricity bills”, the draft NPS repeats (at paragraph 185) that data centre development should not place upward pressure on energy prices. It notes at (paragraph 189) that “Applicants will be expected to make credible commitments for the grid capacity they request and set out the impact that the data centre might have on electricity prices” This is then supplemented in the “decision-making” limb of the policy with various corresponding statements, including that “The Secretary of State must be satisfied that the necessary consideration for securing a grid connection has been given by the applicant. If a connection is not secure at the time of application, the relevant Secretary of State will assess the likelihood that a connection will be secured in enough time for the development to proceed… The Secretary of State should have regard to the extent to which the applicant has a credible plan for supporting the security of supply, investing in clean power or paying for relevant costs of connecting and operating, such that they do not increase bills.”
Interestingly, particularly from a “co-location” and “powered land” perspective, paragraph 191 notes that “Applicants who can demonstrate either a high probability of securing a grid connection, or a reduced demand of the national grid through the use of renewable or low carbon sources, will be given positive weighting in the planning balance by the relevant Secretary of State when making development consent decisions.”
There is no reference to the Local Benefit Plan requirement (likely due to a desire to keep such commitments distinct from the consideration and determination of the planning application in view of the established case-law on such matters), but the draft NPS does seek to incorporate the commitment to requiring adherence to the “highest environmental standards” (paragraph 153 onwards). As part of this, it repeats (at paragraph 155) the direction that – “Applicants should consider opportunities for heat reuse, reflecting its potential contribution to energy efficiency and the supply of low carbon heating via heat networks to local homes and businesses. Where reuse is not pursued, applicants should clearly explain why it is not feasible”. There are obvious practical/commercial challenges with heat reuse (potentially aggravated where development is directed according to regional infrastructure targets), and so applicants will need to carefully consider how this is framed in their application.
The socio-economic and security elements of the Covenant are reflected positively in the Need case (section 4) and within the socio-economic assessment policies (paragraph 391 onwards) and should be more straightforward to evidence by comparison.
Conclusion
The consultation on the draft NPS is open until 8 January 2027. Whilst the consultation responses may add some nuance to the final wording of the policy, it is unlikely to evolve in a way that is inconsistent with the Government’s stated ambitions in the Covenant.
It remains to be seen whether the policy incentives within the Covenant and affirmed within the Connections Consultation Response and draft NPS are sufficient to encourage development and delivery, or whether the practical limitations (particularly if the regional infrastructure targets provide restrictive) and financial obligations prove too much of a deterrent in a globally competitive market.
We would be very happy to discuss strategy and approach with prospective developers who would like to explore the implications further. We are currently advising on the promotion of one of the first data centre schemes to be directed into the NSIP regime, proposed to be developed as part of a co-located scheme with an approximately 1.3-1.4 GW solar generating station that will provide a direct supply of renewable electricity to the approx 1 GW data centre, together with BESS and associated grid infrastructure.8 We would be delighted to discuss standalone data centre development schemes or the potential for co-located schemes and the appropriate consenting routes for either.
- https://www.gov.uk/government/publications/data-centre-covenant/the-data-centre-covenant
- https://assets.publishing.service.gov.uk/media/6ae1fbb8088f42d99bbe781b/accelerating-network-connections-government-response-web.pdf
- https://www.gov.uk/government/consultations/draft-national-policy-statement-for-data-centres
- https://www.reuters.com/world/us/trumps-midterm-drubbing-demonstrates-power-data-centre-opponents-2026-11-04/
- https://www.ofgem.gov.uk/press-release/ofgem-acts-free-grid-capacity-tackling-speculative-data-centre-projects
- https://www.gov.uk/government/publications/energy-independence-bill-factsheets/mandatory-community-benefits-for-generation
- https://questions-statements.parliament.uk/written-statements/detail/2024-09-12/hcws89
- https://longmarshgreenenergy.co.uk/solar-farm-and-data-centre-proposed-in-norfolk
Key contacts
Ian Mack
Partner, London
Catherine Howard
Partner, London
Charlotte Dyer
Of Counsel, London
Naomi Hoggett
Senior Associate, London
Disclaimer
The articles published on this website, current at the dates of publication set out above, are for reference purposes only. They do not constitute legal advice and should not be relied upon as such. Specific legal advice about your specific circumstances should always be sought separately before taking any action.